WO2022213364A1 - Procédé de traitement d'image, appareil de traitement d'image et support de stockage lisible - Google Patents

Procédé de traitement d'image, appareil de traitement d'image et support de stockage lisible Download PDF

Info

Publication number
WO2022213364A1
WO2022213364A1 PCT/CN2021/086195 CN2021086195W WO2022213364A1 WO 2022213364 A1 WO2022213364 A1 WO 2022213364A1 CN 2021086195 W CN2021086195 W CN 2021086195W WO 2022213364 A1 WO2022213364 A1 WO 2022213364A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
target
area
processor
target area
Prior art date
Application number
PCT/CN2021/086195
Other languages
English (en)
Chinese (zh)
Inventor
顾磊
Original Assignee
Oppo广东移动通信有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Oppo广东移动通信有限公司 filed Critical Oppo广东移动通信有限公司
Priority to PCT/CN2021/086195 priority Critical patent/WO2022213364A1/fr
Publication of WO2022213364A1 publication Critical patent/WO2022213364A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/50—Depth or shape recovery
    • G06T7/536—Depth or shape recovery from perspective effects, e.g. by using vanishing points
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/50—Depth or shape recovery
    • G06T7/55—Depth or shape recovery from multiple images
    • G06T7/593—Depth or shape recovery from multiple images from stereo images

Definitions

  • the present application relates to the technical field of image processing, and more particularly, to an image processing method, an image processing apparatus, a terminal, and a non-volatile computer-readable storage medium.
  • an electronic device such as a mobile phone
  • acquires an image of an object and needs to acquire a front view of the object it is necessary to ensure that the camera of the electronic device is facing the object before acquiring the image of the object.
  • the electronic device can only obtain the oblique view of the object, so that after the electronic device finally obtains the image of the object, it is not conducive to obtain the information of the image of the object.
  • Embodiments of the present application provide an image processing method, an image processing apparatus, a terminal, and a non-volatile computer-readable storage medium.
  • the image processing method of the embodiment of the present application includes: collecting a first image of a target object by using a first sensor; collecting depth information of the target information by using the second sensor; determining the target object according to the first image and the depth information. a target image area of the target object in the first image; and performing perspective transformation on the target image area according to the depth information to generate a target image.
  • the image processing apparatus of the embodiment of the present application includes a first acquisition module, a second acquisition module, a processing module, and a generation module.
  • the first acquisition module is used to acquire the first image of the target object through the first sensor.
  • the second acquiring module is configured to acquire the depth information through the second sensor.
  • the processing module is configured to determine a target image area of the target object in the first image according to the first image and the depth information.
  • the generating module is configured to perform perspective transformation on the target image area according to the depth information to generate a target image.
  • a terminal includes a first sensor, a second sensor, and a processor.
  • the first sensor is used for collecting a first image of the target object.
  • the second sensor is used for collecting depth information of the target object.
  • the processor is configured to: determine a target image area of the target object in the first image according to the first image and the depth information; and perform perspective transformation on the target image area according to the depth information, to generate the target image.
  • the non-volatile computer-readable storage medium of the embodiments of the present application includes a computer program, and when the computer program is executed by one or more processors, causes the processors to execute the following image processing method: acquiring a target through a first sensor obtaining a first image of an object; acquiring depth information of the target object through a second sensor; determining a target image area of the target object in the first image according to the first image and the depth information; and The depth information is used to perform perspective transformation on the target image area to generate a target image.
  • the image processing method, image processing device, terminal, and non-volatile computer-readable storage medium of the embodiments of the present application first identify and determine the target image area of the target object in the first image by using the first image and depth information, and then use the depth information to identify and determine the target image area of the target object in the first image.
  • the information determines the perspective relationship of the target object so as to perform perspective transformation on the target image area, so that the generated target image is the front view of the target object, which is beneficial to obtain the information of the target object.
  • FIG. 1 is a schematic flowchart of an image processing method according to some embodiments of the present application.
  • FIG. 2 is a schematic structural diagram of a terminal according to some embodiments of the present application.
  • FIG. 3 is a schematic plan view of an image processing apparatus according to some embodiments of the present application.
  • FIG. 4 is a schematic diagram of a scene of an image processing method according to some embodiments of the present application.
  • 5 to 7 are schematic flowcharts of image processing methods according to some embodiments of the present application.
  • FIG. 8 is a schematic diagram of a scene in which the first target area performs edge transformation according to some embodiments of the present application.
  • FIG. 12 is a schematic diagram of a scene in which the second target area of the present application performs edge transformation
  • FIG. 13 is a schematic flowchart of an image processing method according to some embodiments of the present application.
  • FIGS. 14A to 14D are schematic diagrams of scenes of image processing methods according to some embodiments of the present application.
  • 15 is a schematic flowchart of an image processing method according to some embodiments of the present application.
  • 16 is a schematic diagram of a scene of an image processing method according to some embodiments of the present application.
  • 17 is a schematic diagram of a scene of an image processing method according to some embodiments of the present application.
  • FIG. 18 is a schematic diagram of a connection state between a computer-readable storage medium and a processor according to some embodiments of the present application.
  • an embodiment of the present application provides an image processing method.
  • the image processing method includes the steps:
  • an embodiment of the present application provides an image processing apparatus 10 .
  • the image processing apparatus 10 includes a first acquiring module 11 , a second acquiring module 12 , a processing module 13 and a generating module 14 .
  • the image processing method of the embodiment of the present application can be applied to the image processing apparatus 10 .
  • the first acquisition module 11 is used to perform step 01
  • the second acquisition module 12 is used to perform step 02
  • the processing module 13 is used to perform step 03
  • the generation module 14 is used to perform step 04 . That is, the first acquisition module 11 is configured to acquire the first image of the target object through the first sensor 20 .
  • the second acquisition module 12 is configured to acquire depth information of the target object through the second sensor 30 .
  • the processing module 13 is configured to determine a target image area of the target object in the first image according to the first image and the depth information.
  • the generating module 14 is configured to perform perspective transformation on the target image area according to the depth information, so as to generate the target image.
  • an embodiment of the present application further provides a terminal 100 .
  • the terminal 100 includes a first sensor 20 , a second sensor 30 and a processor 40 .
  • the image processing method of the embodiment of the present application can be applied to the terminal 100 of the embodiment of the present application.
  • the first sensor 20 is used to perform step 01
  • the second sensor 30 is used to perform step 02
  • the processor 40 is used to perform step 03 and step 04 . That is, the first sensor 20 is used to collect the first image of the target object
  • the second sensor 30 is used to collect the depth information of the target object
  • the processor 40 is used to determine the target object in the first image according to the first image and the depth information and performing perspective transformation on the target image area according to the depth information to generate the target image.
  • the number of processors 40 may be one or more. In one example, multiple processors 40 may be integrated inside the first sensor 20 and inside the second sensor 30, respectively, to obtain the first image and depth information, respectively. In another example, when there is one processor 40, part of the processor 40 can be integrated inside the first sensor 20, and another part of the processor 40 can be integrated inside the second sensor 30, so as to acquire the first image and in-depth information.
  • the processor 40 needs to obtain the first image of the target object through the first sensor 20 first, and obtain the depth information of the target object through the second sensor 30 .
  • the first image collected by the first sensor 20 is a two-dimensional image of the target object.
  • the depth information collected by the second sensor 30 is the depth of each position in the target object.
  • the first sensor 20 is a color sensor (RGB sensor)
  • the second sensor 30 is a depth sensor (such as a Time of Flight (TOF), or a structured light sensor)
  • the processor 40 can pass the first sensor 20 After acquiring the two-dimensional image of the target object, the processor 40 may also acquire depth information of the target object through the second sensor 30 and transform the depth information into three-dimensional coordinates in the coordinate system of the second sensor 30 to generate a three-dimensional point cloud.
  • the processor 40 can identify the image area where the target object in the first image is located through the first image, and the processor 40 can perform plane detection on the three-dimensional point cloud, so as to identify the difference between the three-dimensional point cloud and the target object.
  • the corresponding plane thereby extracting the three-dimensional point cloud belonging to the plane.
  • the plane is a plane composed of point clouds with a depth of 5 in a 3D point cloud. Since the target object may have different depths, in order to obtain the point cloud corresponding to the target object, a depth threshold range can be preset.
  • the depth threshold range is 0.1, and the processor 40 can obtain all the three-dimensional point clouds with a depth of 4.9 to 5.1 point cloud.
  • the depth threshold range is an empirical value, which can be determined according to the photographed target object, for example, the depth threshold range is determined according to the thickness of the photographed target object itself.
  • the processor 40 can project the extracted three-dimensional point cloud to the first image, and perform edge point recognition on the extracted three-dimensional point cloud, thereby obtaining a projection image corresponding to the first image collected by the second sensor 30 .
  • the processor 40 may determine the target image area of the target object in the first image by fusing the first image and the projected image. For example, the processor 40 can intercept the intersection of the first image and the projected image to obtain the target image area; the processor 40 can also intercept the union of the first image and the projected image to obtain the target image area.
  • the processor 40 After the processor 40 obtains the target image area, the processor 40 also needs to perform perspective transformation on the target image according to the depth information to generate the target image.
  • the processor 40 when the processor 40 performs perspective transformation on the target image area, the processor 40 first calculates the coordinates of the transformed vertices of the target image area, thereby determining the minimum circumscribed rectangle of the transformed target image area. At this time, the processor 40 Then, by judging whether the pixels in the rectangle belong to the transformed target image area, the pixels belonging to the transformed target image area can be inversely transformed, so as to obtain the coordinates of each pixel in the first image. Interpolate the nearby pixels (such as linear interpolation according to one or more pixels around each pixel to obtain the pixel value of the pixel) to obtain the pixel value of each pixel in the target image area, thereby obtaining the perspective transformed target image The pixel value of each pixel of the area, and then generate the target image.
  • the nearby pixels such as linear interpolation according to one or more pixels around each pixel to obtain the pixel value of the pixel
  • the left picture of Fig. 4 is the target image region S before the transformation
  • the right picture of Fig. 4 is the transformed target image region S1. Since the area of the transformed target image region S1 is larger than the target image region S before the transformation, Therefore, the processor 40 needs to perform pixel interpolation on the transformed target image area S1 to generate the target image. Taking the M point as an example, since the M point is located in the transformed target image area S1, the processor 40 may perform an inverse transformation on the M point to obtain the coordinates of the M point in the first image, so as to obtain the M point in the first image.
  • the processor 40 can use the average value of the pixel values of the pixels in the M point and the pixels near the M point as the pixel value of the M point in the transformed target image area S1. After all points of the image area S1 are interpolated, the target image can be generated.
  • the terminal 100 may be a mobile phone, a tablet computer, a notebook computer, etc., as well as a terminal 100 having a function of displaying and acquiring images.
  • the terminal 100 is a mobile phone as an example for description. It can be understood that the specific form of the terminal 100 is not limited to a mobile phone.
  • an electronic device such as a mobile phone
  • acquires an image of an object and needs to acquire a front view of the object it is necessary to ensure that the camera of the electronic device is facing the object before acquiring the image of the object.
  • the electronic device can only obtain the oblique view of the object, so that after the electronic device finally obtains the image of the object, it is not conducive to read the information of the image of the object.
  • the first image and the depth information are used to first identify and determine the target image area of the target object in the first image, and then determine the perspective relationship of the target object through the depth information.
  • the perspective transformation is performed on the target image area, so that the generated target image is the front view of the target object, which is beneficial to obtain the information of the target object.
  • step 03 determining the target image area of the target object in the first image according to the first image and the depth information, which may include the steps:
  • the processing module 13 is configured to execute step 031 , step 032 , step 033 and step 034 . That is, the processing module 13 is used to determine the first target area of the target object in the first image according to the first image; generate the depth image according to the depth information; determine the second target area of the target object in the depth image according to the depth information; A target area and a second target area define a target image area.
  • the processor 40 is configured to execute step 031 , step 032 , step 033 and step 034 . That is, the processor 40 is configured to determine the first target area of the target object in the first image according to the first image; generate the depth image according to the depth information; determine the second target area of the target object in the depth image according to the depth information; A target area and a second target area define a target image area.
  • the first image is an image including the target object
  • the depth information is depth information including the target object.
  • the processor 40 may identify and cut out the position of the target object in the first image according to the first image, that is, the two-dimensional image of the target object, so as to obtain the first target area.
  • the processor 40 can use the depth information to generate a depth image of the target object, and the processor 40 can identify the target object in the three-dimensional point cloud according to the three-dimensional point cloud generated by the second sensor 30 depth information corresponding to the first image of the object, so as to determine the second target area of the target object in the depth image according to the depth information.
  • the processor 40 may determine the target image area by fusing the first target area and the second target area. For example, the processor 40 can intercept the intersection of the first target area and the second target area to obtain the target image area; the processor 40 can also intercept the union of the first target area and the second target area to obtain the target image area.
  • step 031 determining the first target area of the target object in the first image according to the first image, which may include steps:
  • 0312 Perform contour detection on the grayscale image to determine the first target area.
  • the processing module 13 is used to perform steps 0311 and 0312, that is, the processing module 13 is used to convert the first image into a grayscale image; perform contour detection on the grayscale image to determine the first image. a target area.
  • the processor 40 is used to perform steps 0311 and 0312, that is, the processor 40 is used to convert the first image into a grayscale image; perform contour detection on the grayscale image to determine the first image. a target area.
  • the processor 40 may convert the first image into a grayscale image, so as to determine the first target area by performing contour detection on the grayscale image.
  • the processor 40 is provided with a predetermined grayscale threshold, and the processor 40 can compare the grayscale value of each pixel and the average grayscale value of all pixels in the grayscale image to determine whether the grayscale value exceeds the grayscale threshold range.
  • the image is subjected to contour detection to determine the first target image. For example, if the grayscale threshold range is 1 and the average grayscale value of the grayscale image is 150, when the difference between the grayscale value of some pixels in the grayscale image and the average grayscale value is between -1 and 1,
  • the processor 40 determines that the part of the pixels is the pixel corresponding to the target object in the first image, and the outline of the figure formed by the part of the pixels is the outline of the first target area, thereby determining the first target area.
  • the gray threshold range is an empirical value, which can be determined according to the photographed target object, for example, the depth threshold range is determined according to the color of the photographed target object itself.
  • the processor 40 may also convert the grayscale image into a binary image to perform contour detection on the grayscale image to determine the first target area. At this time, the processor 40 may set multiple grayscale threshold ranges according to different target objects acquired by the first sensor 20 .
  • the processor 40 needs to select a larger grayscale threshold. If the grayscale threshold is 10, the average grayscale value of the grayscale image is 150, then when the absolute value of the difference between the grayscale value of the pixel in the grayscale image and the average grayscale value is greater than 10, the processor 40 determines that the pixel is the pixel of the contour of the target object, and denote the pixel as 1, Then, after the processor 40 determines that all pixels are completed, the figure formed by the position displayed as 1 in the binary image is the outline of the first target area, thereby determining the first target area.
  • the processor 40 needs to select a smaller grayscale threshold. If the grayscale threshold is 1, the average grayscale value of the grayscale image is is 150, then when the absolute value of the difference between the gray value of the pixel in the first image and the average gray value is greater than 1, the processor 40 determines that the pixel is the pixel of the contour of the target object, and denote the pixel as 1 , then when the processor 40 determines that all pixels are completed, the figure formed by the position displayed as 1 in the binary image is the outline of the first target area, thereby determining the first target area.
  • the processor 40 may further perform filtering processing on the grayscale image, so as to filter out the pixels in the grayscale image that do not belong to the part of the target object in advance, Thus, the grayscale image is quickly detected, thereby determining the first target area.
  • the processor 40 can also identify the density of the pixels in the grayscale image, because when the first sensor 20 acquires the first image containing the target object, the pixel distribution of the part of the first image corresponding to the target object is compared with The pixel distribution of the first image of the part of the non-target object is different, and the processor 40 can judge whether the density of the pixels in the edge area is consistent with the density of the pixels in the center area in all the pixels in the grayscale image. Whether the pixel is a pixel of the first image corresponding to the portion of the target object.
  • the processor 40 can filter out the part of the pixels. After the processor 40 has filtered out all the pixels, the outline of the grayscale image at this time is the outline of the first target area, thereby determining the first target area.
  • step 0312 performing contour detection on the grayscale image to determine the first target area, including the steps of:
  • the processing module 13 is used to execute step 0313, that is, the processing module 13 is used to divide the two sides of the first target area when the angle between the two adjacent sides is greater than the predetermined angle. Convert to an edge.
  • the processor 40 is configured to perform step 0313, that is, the processor 40 is configured to divide the two sides of the first target area when the angle between the two adjacent sides is greater than the predetermined angle. Convert to an edge.
  • the processor 40 In order to optimize the shape of the first target area, after the processor 40 performs contour detection on the grayscale image to determine the first target area, the processor 40 also needs to optimize the edges of the first target area, so as to obtain a relatively regular first target area. a target area.
  • a predetermined angle may be set in the processor 40, and after the processor 40 determines the first target area, the processor 40 may determine whether the angle between any adjacent two sides of the first target area is greater than the predetermined angle, and When the included angle between any adjacent two sides of the first target area is greater than the predetermined included angle, the two sides are converted into one side.
  • the processor 40 converts L1 and L2 into one side L3, Thus, the contour-optimized first target area 50 is obtained.
  • the predetermined included angle may also be the value of the third included angle after the included angles of all adjacent two sides in the first target area 50 are sorted in descending order. For example, when the processor 40 obtains that the included angles of all adjacent two sides in the first target area 50 are 90 degrees, 190 degrees, 120 degrees, 140 degrees, 150 degrees, 110 degrees, and 100 degrees, respectively, the processor 40 determines these 7 The included angles are sorted from small to large as 190 degrees, 150 degrees, 140 degrees, 120 degrees, 110 degrees, 100 degrees and 90 degrees. Then the predetermined included angle is 140 degrees, the processor 40 converts the two sides with an included angle of 190 degrees into one side, and the processor 40 also converts the two sides with an included angle of 150 degrees into one side. Wherein, if two sides with an included angle of 190 degrees and two sides with an included angle of 150 degrees share one side, the processor 40 will convert the three sides into one side.
  • step 033 determining the second target area of the target object in the depth image according to the depth information, including the steps:
  • 0331 Identify pixels within the depth range of the target object in the depth image based on a predetermined algorithm to generate a second target area.
  • the processing module 13 is configured to perform step 0331 , that is, the processing module 13 is configured to identify pixels within the depth range of the target object in the depth image based on a predetermined algorithm to generate a second target area.
  • the processor 40 is configured to perform step 0331 , that is, the processor 40 is configured to identify pixels within the depth range of the target object in the depth image based on a predetermined algorithm to generate a second target area.
  • the processor 40 can obtain a three-dimensional image of the target object according to the depth image.
  • the processor 40 can perform plane detection on the three-dimensional image to identify the plane corresponding to the target object in the three-dimensional image, so as to extract the pixels belonging to the plane.
  • the processor 40 is set with a predetermined depth range, and after the processor 40 obtains a plane corresponding to the first target area in the three-dimensional image, the processor can extract pixels belonging to the plane.
  • the plane is a plane composed of pixels with a depth of 10 in the 3D image. Since the target object may have different depths, in order to obtain the pixels corresponding to the target object, the depth threshold range can be preset. For example, the depth threshold range is 1.
  • the processor 40 may, among all pixels, project the pixels with a depth of 9 to 11 on the first target area, and perform edge point recognition on the extracted pixels, thereby generating a second target area.
  • the processor 40 may further perform filtering processing on all the acquired pixels, so as to filter out the pixels that do not belong to the target object among all the pixels in advance, so as to ensure the generation of the first target area. 2. The accuracy of the target area.
  • the processor 40 can also pass the density of all pixels, because when the second sensor 30 acquires the depth information containing the target object, the pixel distribution of the depth image of the part corresponding to the target object is compared with the depth of the part of the non-target object The pixel distributions of the images are different, and the processor 40 can determine whether the pixel is part of the corresponding target object by judging whether the density of the pixels in the edge area is consistent with the density of the pixels in the center area in all the pixels in the three-dimensional image. Pixels of the depth image.
  • the processor 40 can filter out the part of the pixels. After the processor 40 filters out all the pixels, the pixels in the three-dimensional image at this time are the pixels of the depth image corresponding to the target object, and the processor 40 can accurately generate the second target area.
  • step 034 determining a target image area according to the first target area and the second target area, including steps:
  • the processing module 13 is configured to perform steps 0341 and 0342 , that is, the processing module 13 is configured to convert the second target area into the second target area based on the position conversion relationship between the first sensor 20 and the second sensor 30 mapping into the first image to generate a third target area; and determining a target image area according to the first target area and the third target area.
  • the processor 40 is configured to perform steps 0341 and 0342 , that is, the processor 40 is configured to convert the second target area into the second target area based on the position conversion relationship between the first sensor 20 and the second sensor 30 mapping into the first image to generate a third target area; and determining a target image area according to the first target area and the third target area.
  • the processor 40 needs to map the second target area to the first image, and convert the three-dimensional image of the target object into a two-dimensional image. Each point is established according to the coordinate system of the first sensor 20 , and the depth information is established according to the coordinate system of the second sensor 30 , then the processor 40 needs to convert the position of the first sensor 20 and the second sensor 30 into Each point on the two target areas is converted into a new second target area established with the coordinate system of the first sensor 20 .
  • the origin of the coordinate system of the first sensor 20 can be moved by 3 units in the positive direction of the X axis to coincide with the origin of the coordinate system of the second sensor 30, then when the second target area is mapped to the first image, the second target area The X-coordinates of all points of , need to be reduced by 3 units to generate the third target area.
  • the processor 40 may determine the target image area by fusing the first target area and the third target area. For example, the processor 40 can intercept the intersection of the first target area and the third target area to obtain the target image area; the processor 40 can also intercept the union of the first target area and the third target area to obtain the target image area.
  • step 0341 map the second target area to the first image based on the position conversion relationship between the first sensor 20 and the second sensor 30 to generate a third target area, including step:
  • the processing module 13 is used to execute step 0343, that is, the processing module 13 is used to convert the two sides when the angle between any adjacent two sides of the third target area is greater than the predetermined angle. for a side.
  • the processor 40 is used to perform step 0343, that is, the processor 40 is used to convert the two sides when the angle between any adjacent two sides of the third target area is greater than the predetermined angle. for a side.
  • the processor 40 In order to optimize the shape of the third target area, after the processor 40 determines the third target area, the processor 40 further needs to perform optimization processing on the edges of the third target area, so as to obtain a relatively regular third target area.
  • a predetermined angle may be set in the processor 40, and after the processor 40 determines the third target area, the processor 40 may determine whether the angle between any adjacent two sides of the third target area is greater than the predetermined angle, and When the included angle between any adjacent two sides of the third target area is greater than the predetermined included angle, the two sides are converted into one side.
  • the processor 40 converts P1 and P2 into a side P3,
  • the third target area 60 with the optimized contour is obtained.
  • the predetermined included angle may also be the value of the second included angle after the included angles of all adjacent two sides in the third target area 60 are sorted in descending order. For example, when the processor 40 obtains that the included angles of all adjacent two sides in the third target area 60 are 50 degrees, 140 degrees, 120 degrees, 130 degrees, and 100 degrees, respectively, the processor 40 measures the five included angles from small to large. Sorting is 140 degrees, 130 degrees, 120 degrees, 100 degrees, and 50 degrees. Then the predetermined included angle is 130 degrees, and the processor 40 converts the two sides whose included angle is 140 degrees into one side.
  • step 0342 determining the target image area according to the first target area and the third target area, including steps:
  • 0344 Determine the target image area according to the intersection, union, envelope of the union, or expansion of the intersection between the first target area and the third target area.
  • the processing module 13 is configured to perform step 0344, that is, the processing module 13 is configured to perform an envelope or intersection according to the intersection, union, and union of the first target area and the third target area.
  • the dilation determines the target image area.
  • the processor 40 is configured to perform step 0344, that is, the processor 40 is configured to perform an envelope or intersection according to the intersection, union, and union of the first target area and the third target area.
  • the dilation determines the target image area.
  • the processor 40 may determine the target image area according to the intersection, union, and envelope or intersection of the first target area and the third target area of dilation to determine the target image area.
  • the envelope of the union is the smallest circumscribed rectangle containing the first target area and the third target area, and the expansion of the intersection is an enlarged image of the intersection of the first target area and the third target area.
  • the processor 40 can use the first target area 70 and the third target area 60 to The intersection, the union, the envelope of the union, or the expansion of the intersection are obtained, respectively, of the first target region 70 and the third target region 60 . Since only one image showing the most accurate target object needs to be finally obtained, the processor 40 may identify the intersection, union, envelope of the union or the expansion of the intersection, respectively, to obtain the One image of the first image can be completely displayed as the target image area.
  • the processor 40 uses the intersection of the first target area 70 and the third target area 60 as the target image area.
  • the processor 40 takes the intersection of the first target area 70 and the third target area 60 as the target image area.
  • step 04 performing perspective transformation on the target image area according to the depth information, before, it also includes steps:
  • the generation module 14 is used to perform steps 05 and 06, that is, the generation module 14 is used to display the first image and the target image area in the preview image; receive input instructions to adjust the target image the boundary of the area.
  • the processor 40 is used to perform steps 05 and 06, that is, the processor 40 is used to display the first image and the target image area in the preview image; receive input instructions to adjust the target image the boundary of the area.
  • the processor 40 outputs the preview image to the terminal 100 for display, so as to The user manually corrects the target image area so that the target image area completely matches the first image, and the processor 40 can receive the correction instruction input by the user, so as to adjust the boundary of the target image area so that the boundary of the target image area matches the first image.
  • the boundaries of an image match exactly.
  • the target image area 80 obtained by the processor 40 can completely display the first image 90
  • the target image area 80 is 4-sided
  • the first image 90 is 6-sided shape
  • the target image area 80 has only 4 adjustment points
  • the first image 90 has 6 vertices.
  • the processor 40 can add an adjustment point at the midpoint of each edge of the target image area 80 to make the target image
  • the number of adjustment points in the area 80 is greater than the vertices of the first image 90
  • the user can drag the positions of the six appropriate adjustment points in the target image area 80 to make the target image area 80 coincide with the first image 90, thereby
  • the target image area 80 is made to exactly match the first image 90 .
  • the generation module 14 is used to perform steps 07 and 08, that is, the generation module 14 is used to determine whether the terminal 100 is in a stable state; Steps for image and target image area.
  • the processor 40 is used to perform steps 07 and 08, that is, the processor 40 is used to determine whether the terminal 100 is in a stable state; Steps for image and target image area.
  • the terminal 100 may further include a stability detector (not shown in the figure), and the stability detector is used to determine whether the terminal 100 is in a stable state.
  • the stability detector may determine whether the terminal 100 is in a stable state according to the first image and depth information collected by the first sensor 20 and the second sensor 30 .
  • the stability detector may determine whether the terminal 100 is in a stable state according to a gyroscope and a gravity sensor (not shown) of the terminal 100 .
  • the stability detector determines that the terminal 100 is in a stable state
  • the stability detector transmits the determination result to the processor 40
  • the processor 40 determines that the first image and depth information acquired by the first sensor 20 and the second sensor 30 are valid data
  • the processor 40 will output the preview image to the terminal 100 for display, so that the user can manually correct the target image area so that the target image area completely matches the first image, and the processor 40
  • the correction instruction input by the user can be received, so as to adjust the boundary of the target image area, so that the boundary of the target image area completely matches the boundary of the first image.
  • the stability detector determines that the terminal 100 is not in a stable state
  • the stability detector transmits the determination result to the processor 40
  • the processor 40 determines that the first image and depth information acquired by the first sensor 20 and the second sensor 30 are invalid data
  • the processor 40 may not process the first image and depth information acquired by the first sensor 20 and the second sensor 30, and output invalid data to the terminal 100 for display, reminding the user to re-shoot the target object.
  • an embodiment of the present application further provides a non-volatile computer-readable storage medium 200 containing a computer program 201.
  • the computer program 201 When executed by one or more processors 40, the computer program 201 causes the one or more processors 40 to execute the image processing method of any one of the above-described embodiments.
  • the computer program 201 when executed by one or more processors 40, causes the processors 40 to perform the following image processing methods:
  • the computer program 201 when executed by one or more processors 40, it causes the processors 40 to execute the following image processing methods:
  • the computer program 201 when executed by one or more processors 40, it causes the processors 40 to execute the following image processing methods:
  • 0312 Perform contour detection on the grayscale image to determine the first target area.
  • the processors 40 are caused to execute the following image processing methods:
  • the computer program 201 when executed by one or more processors 40, causes the processors 40 to perform the following image processing methods:
  • 0331 Identify pixels within the depth range of the target object in the depth image based on a predetermined algorithm to generate a second target area.
  • first and second are only used for descriptive purposes, and should not be construed as indicating or implying relative importance or implying the number of indicated technical features. Thus, a feature delimited with “first”, “second” may expressly or implicitly include at least one of that feature.
  • plurality means at least two, such as two, three, etc., unless expressly and specifically defined otherwise.
  • any description of a process or method in the flowcharts or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing a specified logical function or step of the process , and the scope of the preferred embodiments of the present application includes alternative implementations in which the functions may be performed out of the order shown or discussed, including performing the functions substantially concurrently or in the reverse order depending upon the functions involved, which should It is understood by those skilled in the art to which the embodiments of the present application belong.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

L'invention concerne un procédé de traitement d'image, un appareil de traitement d'image (10), un terminal (100), ainsi qu'un support de stockage non volatil lisible par ordinateur (200). Le procédé de traitement d'image consiste à : acquérir une première image d'un objet cible à l'aide d'un premier capteur (20) ; acquérir des informations de profondeur de l'objet cible à l'aide d'un second capteur (30) ; déterminer une zone d'image cible en fonction de la première image et des informations de profondeur ; et effectuer une transformation de perspective sur la zone d'image cible selon les informations de profondeur afin de générer une image cible.
PCT/CN2021/086195 2021-04-09 2021-04-09 Procédé de traitement d'image, appareil de traitement d'image et support de stockage lisible WO2022213364A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/086195 WO2022213364A1 (fr) 2021-04-09 2021-04-09 Procédé de traitement d'image, appareil de traitement d'image et support de stockage lisible

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/086195 WO2022213364A1 (fr) 2021-04-09 2021-04-09 Procédé de traitement d'image, appareil de traitement d'image et support de stockage lisible

Publications (1)

Publication Number Publication Date
WO2022213364A1 true WO2022213364A1 (fr) 2022-10-13

Family

ID=83545073

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/086195 WO2022213364A1 (fr) 2021-04-09 2021-04-09 Procédé de traitement d'image, appareil de traitement d'image et support de stockage lisible

Country Status (1)

Country Link
WO (1) WO2022213364A1 (fr)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190012796A1 (en) * 2015-09-10 2019-01-10 Sony Corporation Image processing apparatus and method
CN110335224A (zh) * 2019-07-05 2019-10-15 腾讯科技（深圳）有限公司 图像处理方法、装置、计算机设备及存储介质
CN111179332A (zh) * 2018-11-09 2020-05-19 北京市商汤科技开发有限公司 图像处理方法及装置、电子设备及存储介质
WO2020237611A1 (fr) * 2019-05-31 2020-12-03 深圳市大疆创新科技有限公司 Procédé et appareil de traitement d'image, terminal de commande et dispositif mobile
CN112085775A (zh) * 2020-09-17 2020-12-15 北京字节跳动网络技术有限公司 图像处理的方法、装置、终端和存储介质
CN112581629A (zh) * 2020-12-09 2021-03-30 中国科学院深圳先进技术研究院 增强现实显示方法、装置、电子设备及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190012796A1 (en) * 2015-09-10 2019-01-10 Sony Corporation Image processing apparatus and method
CN111179332A (zh) * 2018-11-09 2020-05-19 北京市商汤科技开发有限公司 图像处理方法及装置、电子设备及存储介质
WO2020237611A1 (fr) * 2019-05-31 2020-12-03 深圳市大疆创新科技有限公司 Procédé et appareil de traitement d'image, terminal de commande et dispositif mobile
CN110335224A (zh) * 2019-07-05 2019-10-15 腾讯科技（深圳）有限公司 图像处理方法、装置、计算机设备及存储介质
CN112085775A (zh) * 2020-09-17 2020-12-15 北京字节跳动网络技术有限公司 图像处理的方法、装置、终端和存储介质
CN112581629A (zh) * 2020-12-09 2021-03-30 中国科学院深圳先进技术研究院 增强现实显示方法、装置、电子设备及存储介质

Similar Documents

Publication Publication Date Title
US11468579B2 (en) Method and device for image processing
CN109872397B (zh) 一种基于多目立体视觉的飞机零件的三维重建方法
US20200058153A1 (en) Methods and Devices for Acquiring 3D Face, and Computer Readable Storage Media
JP6125188B2 (ja) 映像処理方法及び装置
US20180218507A1 (en) 3d reconstruction of a human ear from a point cloud
JP6955783B2 (ja) 情報処理方法、装置、クラウド処理デバイス及びコンピュータプログラム製品
US11042966B2 (en) Method, electronic device, and storage medium for obtaining depth image
US20170140537A1 (en) System and method for scoring clutter for use in 3d point cloud matching in a vision system
JP6073858B2 (ja) 顔の位置検出
TW201214335A (en) Method and arrangement for multi-camera calibration
US9147279B1 (en) Systems and methods for merging textures
CN109064533B (zh) 一种3d漫游方法及系统
CN114981845A (zh) 图像扫描方法及装置、设备、存储介质
WO2020151078A1 (fr) Procédé et appareil de reconstruction en trois dimensions
JP6579659B2 (ja) 光源推定装置及びプログラム
US20230290061A1 (en) Efficient texture mapping of a 3-d mesh
JP7298687B2 (ja) 物体認識装置及び物体認識方法
WO2022213364A1 (fr) Procédé de traitement d'image, appareil de traitement d'image et support de stockage lisible
CN116402904A (zh) 一种基于激光雷达间和单目相机的联合标定方法
CN116051736A (zh) 一种三维重建方法、装置、边缘设备和存储介质
US10339702B2 (en) Method for improving occluded edge quality in augmented reality based on depth camera
US11699303B2 (en) System and method of acquiring coordinates of pupil center point
WO2019016879A1 (fr) Dispositif et procédé de détection d'objet
CN115836322A (zh) 图像裁剪方法与装置、电子设备及存储介质
CN111630569A (zh) 双目匹配的方法、视觉成像装置及具有存储功能的装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21935591

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21935591

Country of ref document: EP

Kind code of ref document: A1